Cognitive decline is a major public health problem that may disproportionately affect older Latinos compared to other older Americans. Growing evidence from epidemiological studies suggests that physical activity (PA), a modifiable health behavior, may improve cognitive function and reduce cognitive decline in older adults. Older Latinos have not been targeted for inclusion in studies of PA and cognitive decline, despite their growing numbers, their increased risk of cognitive impairment, and their relatively poor overall health status. The purpose of the proposed study is to explore the effect of lifestyle PA on cognitive decline in older Latinos. The study will build on the Cognition and Physical Activity Community Elderly Study (CAPACES) to collect a second time-point of PA data and cognitive function. CAPACES is a descriptive, cross-sectional study of 174 Latinos who were 50 years of age or older and without ambulatory disabilities or dementia at baseline, and recruited from 11 predominantly Latino communities in the Chicago metropolitan area. The proposed study aims are: (1) examine the effects of lifestyle PA, as measured with self-report and accelerometer at baseline, on change in cognitive function measured five to six years later in older Latinos, controlling for background characteristics (demographics, acculturation, previous health status); and (2) examine the mediating effects of change in PA on the relation between baseline background characteristics and change in cognitive function. PA behavior will be reassessed subjectively by self-report using the Community Healthy Activities Model Program for Seniors (CHAMPS) for leisure and household PA and objectively by the Actigraph accelerometer, which captures PA counts, duration, and intensity. Cognitive function will be reassessed with a battery of five neurocognitive tests of executive function and episodic memory, the cognitive abilities most frequently shown to be influenced by PA. Additional measures include acculturation, history of diabetes and hypertension, and Center for Epidemiological Studies-Depression scale. A hierarchal regression model will be used to regress change in the five cognitive function measures (outcome variables) on baseline PA controlling for background characteristics. A mediation model analysis will be conducted, made up of background characteristics, change in PA, and change in cognitive function over five to six years. The study findings will inform the development of interventions for older Latinos that increase lifestyle PA to maintain cognitive function with age.